Compare commits

..
2 Commits
Author SHA1 Message Date
sneak 9bc631315d SPEC: the request log is not one line per request
check / check (push) Waiting to run
The package list said requestlog writes one JSON line per request. SPEC's
"The log" section gives a request waf answers three request log lines.

Model: opus-5-5
2026-11-19 17:59:50 +00:00
sneak 1002433c35 Constructors take Params: ratelimit.New and processlog.New (closes #450)
check / check (push) Waiting to run
ratelimit.New takes ratelimit.Params (Limits, MaxClients) in place of
two arguments by position. The process logger moves out of requestlog
into internal/processlog, whose New takes Params (Output, InstanceName,
Level); its two tests move with it unchanged. geojs.GeoJS keeps its
params whole, as DNSBL, AbuseIPDB and routewatch.Client do. SPEC and
README list the new package.

Model: opus-5-5
2026-11-19 17:38:09 +00:00
24 changed files with 404 additions and 635 deletions

No files matched your search

+5 -3
View File
@@ -906,9 +906,11 @@ stops the running process (see [`docs/state-files.md`](docs/state-files.md)).
`internal/coreruleset`, `internal/geojs`, `internal/lookup`,
`internal/reputation`, `internal/routewatch`, `internal/anomaly` and
`internal/alerts`: one check or source each.
- `internal/state`, `internal/requestlog`, `internal/remotelog` and
`internal/metrics`: the state and cache files, the log lines, the syslog
sender and the metrics.
- `internal/state`: the state and cache files.
- `internal/requestlog`: the request log.
- `internal/processlog`: the process's own messages.
- `internal/remotelog`: the syslog sender.
- `internal/metrics`: the metrics.
## Rationale
+4 -2
View File
@@ -104,8 +104,10 @@ never as a time.
its own.
- `internal/state`: reads the state and cache files at start, takes in edits
while running, and writes the files when due and at the stop.
- `internal/requestlog` and `internal/remotelog`: the lines on stdout, and
sending them to the syslog server.
- `internal/requestlog`: the request log, as JSON lines on stdout.
- `internal/processlog`: the process lines on stdout, which hold the
process's own messages.
- `internal/remotelog`: sends the lines on stdout to the syslog server.
- `internal/metrics`: the metrics, in the Prometheus text format.
- `Dockerfile`: the lint and test phases, then the image.
`deploy/example-app` is an app built on the image, with `rproxy` in its
+27 -70
View File
@@ -84,12 +84,7 @@ type Params struct {
// one request to GeoJS is under way at a time, and it asks about every
// client waiting, up to maxPerRequest. It is safe for concurrent use.
type GeoJS struct {
url string
answered func(Answer)
now func() time.Time
asked func()
failed func(err error, askingAgainIn time.Duration)
wentWithoutAnswer func()
params Params
// httpClient follows no redirect, so that visitors' addresses go to
// GeoJS alone: a redirect is a failure.
httpClient *http.Client
@@ -99,13 +94,8 @@ type GeoJS struct {
// waiting are the clients without an answer: those to ask GeoJS about,
// and those it is being asked about.
waiting map[netip.Prefix]*wait
// asking is true while a request to GeoJS is under way, and endAsking
// ends it. stopped is true once StopAsking has been called, and
// askGoroutines waits for the goroutine that makes the requests.
asking bool
endAsking context.CancelFunc
stopped bool
askGoroutines sync.WaitGroup
// asking is true while a request to GeoJS is under way.
asking bool
// retryDelay is how long GeoJS is left alone after its last failure,
// zero after an answer; retryAt is when it may be asked again.
retryDelay time.Duration
@@ -145,12 +135,7 @@ func New(params Params) *GeoJS {
}
return &GeoJS{
url: params.URL,
answered: params.Answered,
now: params.Now,
asked: params.Asked,
failed: params.Failed,
wentWithoutAnswer: params.WentWithoutAnswer,
params: params,
httpClient: &http.Client{
CheckRedirect: func(*http.Request, []*http.Request) error {
return http.ErrUseLastResponse
@@ -193,7 +178,7 @@ func (g *GeoJS) LookUp(
answer, found := g.kept(client)
if !found {
g.wentWithoutAnswer()
g.params.WentWithoutAnswer()
}
w, waiting := g.waiting[client]
@@ -215,23 +200,6 @@ func (g *GeoJS) LookUpWithoutWaiting(
return answer
}
// StopAsking ends the request to GeoJS under way, and returns once the
// goroutine that made it has. A request it ends is no failure: Failed is not
// called, and GeoJS is not left alone after it. After StopAsking, GeoJS is
// asked nothing.
func (g *GeoJS) StopAsking() {
g.mu.Lock()
g.stopped = true
if g.asking {
g.endAsking()
}
g.mu.Unlock()
g.askGoroutines.Wait()
}
// Kept returns client's answer, if one is kept, without asking GeoJS.
func (g *GeoJS) Kept(client netip.Prefix) (Answer, bool) {
g.mu.Lock()
@@ -273,7 +241,7 @@ func (g *GeoJS) Load(answers []Answer) {
g.answers.Purge()
now := g.now()
now := g.params.Now()
for _, answer := range answers {
if now.Sub(answer.Answered) < keepFor {
@@ -310,19 +278,18 @@ func (g *GeoJS) answerOrWait(
}
if w == nil {
g.wentWithoutAnswer()
g.params.WentWithoutAnswer()
return Answer{}, nil // too many clients wait already
}
if !g.asking {
// GeoJS is left alone after a failure, or asked nothing after
// StopAsking, so no answer can come.
// GeoJS is left alone after a failure, so no answer can come.
w.late = true
}
if w.late {
g.wentWithoutAnswer()
g.params.WentWithoutAnswer()
return Answer{}, nil
}
@@ -333,7 +300,7 @@ func (g *GeoJS) answerOrWait(
// kept returns client's answer, if GeoJS gave it less than keepFor ago,
// and notes that it was used.
func (g *GeoJS) kept(client netip.Prefix) (Answer, bool) {
now := g.now()
now := g.params.Now()
kept, found := g.answers.Get(client)
if !found || now.Sub(kept.Answered) >= keepFor {
@@ -346,28 +313,22 @@ func (g *GeoJS) kept(client netip.Prefix) (Answer, bool) {
}
// ask starts asking GeoJS about the waiting clients, unless a request to
// it is under way, it is left alone after a failure, or StopAsking has
// been called. The requests to GeoJS are for every client waiting, so they
// go on when the caller whose ctx is given stops waiting, until StopAsking.
// it is under way or it is left alone after a failure. The requests to
// GeoJS are for every client waiting, so they go on when the caller whose
// ctx is given stops waiting.
func (g *GeoJS) ask(ctx context.Context) {
if g.asking || g.stopped || g.now().Before(g.retryAt) {
if g.asking || g.params.Now().Before(g.retryAt) {
return
}
askCtx, end := context.WithCancel(context.WithoutCancel(ctx))
g.asking, g.endAsking = true, end
g.asking = true
g.askGoroutines.Go(func() {
defer end()
g.askAboutWaiting(askCtx)
})
go g.askAboutWaiting(context.WithoutCancel(ctx))
}
// askAboutWaiting asks GeoJS about the waiting clients, one request at a
// time, until none is left, GeoJS fails, or StopAsking ends ctx. Each
// answer kept is given to Answered, outside the lock, since Answered takes
// locks of its own.
// time, until none is left or GeoJS fails. Each answer kept is given to
// Answered, outside the lock, since Answered takes locks of its own.
func (g *GeoJS) askAboutWaiting(ctx context.Context) {
for {
clients := g.nextClients()
@@ -376,15 +337,11 @@ func (g *GeoJS) askAboutWaiting(ctx context.Context) {
}
given, err := g.request(ctx, clients)
if err != nil && ctx.Err() != nil {
continue // ended by StopAsking, which is no failure
}
kept, answered := g.keep(clients, given, err)
if g.answered != nil {
if g.params.Answered != nil {
for _, answer := range kept {
g.answered(answer)
g.params.Answered(answer)
}
}
@@ -395,13 +352,13 @@ func (g *GeoJS) askAboutWaiting(ctx context.Context) {
}
// nextClients returns up to maxPerRequest of the waiting clients. When
// none is waiting, or StopAsking has been called, it returns none and
// notes that no request to GeoJS is under way.
// none is waiting, it returns none and notes that no request to GeoJS is
// under way.
func (g *GeoJS) nextClients() []netip.Prefix {
g.mu.Lock()
defer g.mu.Unlock()
if len(g.waiting) == 0 || g.stopped {
if len(g.waiting) == 0 {
g.asking = false
return nil
@@ -433,7 +390,7 @@ func (g *GeoJS) keep(
g.mu.Lock()
defer g.mu.Unlock()
now := g.now()
now := g.params.Now()
kept := make([]Answer, 0, len(clients))
leftOut := 0
@@ -477,7 +434,7 @@ func (g *GeoJS) keep(
w.late = true
}
g.failed(err, g.retryDelay)
g.params.Failed(err, g.retryDelay)
return kept, false
}
@@ -507,7 +464,7 @@ func (g *GeoJS) request(
req, err := http.NewRequestWithContext(
ctx,
http.MethodGet,
g.url,
g.params.URL,
http.NoBody,
)
if err != nil {
@@ -516,7 +473,7 @@ func (g *GeoJS) request(
req.URL.RawQuery = "ip=" + strings.Join(addrs, ",")
g.asked()
g.params.Asked()
res, err := g.httpClient.Do(req)
if err != nil {
-38
View File
@@ -318,44 +318,6 @@ func TestCallerWhoseContextEndsEarlyDoesNotStopTheRequest(t *testing.T) {
})
}
func TestStopAskingEndsTheRequestUnderWayAsNoFailureAndAsksNoMore(
t *testing.T,
) {
t.Parallel()
synctest.Test(t, func(t *testing.T) {
service, _, g, reports := startWithReports()
clients := newClients()
service.set(hanging)
g.LookUpWithoutWaiting(t.Context(), clients())
waitForRequests(t, service, 1)
// The request is ended, not waited out, and is no failure.
began := time.Now()
g.StopAsking()
if waited := time.Since(began); waited != 0 {
t.Errorf("StopAsking returned after %s, want at once", waited)
}
reports.wantFailures(t)
// A new client is then not asked about, and its lookup does not
// wait.
service.set(answering)
wantCountry(t, g, clients(), "")
if waited := time.Since(began); waited != 0 {
t.Errorf("the lookup returned after %s, want at once", waited)
}
waitForRequests(t, service, 1)
reports.wantFailures(t)
})
}
func TestRequestOver10SecondsIsAFailure(t *testing.T) {
t.Parallel()
+44
View File
@@ -0,0 +1,44 @@
// Package processlog builds the logger for the process's own messages,
// which it writes as JSON lines marked "type":"process". The request log
// is package requestlog's.
package processlog
import (
"io"
"log/slog"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
)
// Params are what New needs.
type Params struct {
// Output is where the lines are written.
Output io.Writer
// InstanceName, SWW_WAF_INSTANCE_NAME, is each line's instance.
InstanceName string
// Level, SWW_WAF_LOG_LEVEL, is the least severe level written.
Level slog.Level
}
// New returns the logger for the process's own messages: JSON lines on
// params.Output, marked "type":"process", with the time in the same form
// as a request line's. It writes only the messages at params.Level or more
// severe; the request lines requestlog.Write writes are never held back.
func New(params Params) *slog.Logger {
handler := slog.NewJSONHandler(params.Output, &slog.HandlerOptions{
Level: params.Level,
ReplaceAttr: func(groups []string, attr slog.Attr) slog.Attr {
if attr.Key == slog.TimeKey && len(groups) == 0 {
return slog.String(
slog.TimeKey,
requestlog.FormatTime(attr.Value.Time()),
)
}
return attr
},
})
return slog.New(handler).
With("type", "process", "instance", params.InstanceName)
}
+105
View File
@@ -0,0 +1,105 @@
package processlog_test
import (
"bytes"
"encoding/json"
"log/slog"
"slices"
"strings"
"testing"
"time"
"sneak.berlin/go/smallwebwaf/internal/processlog"
)
// instanceName is the instance the tests' loggers are built with.
const instanceName = "host1/blog"
func TestProcessLinesAreMarkedProcessAndGiveTheInstance(t *testing.T) {
t.Parallel()
var out bytes.Buffer
processlog.New(processlog.Params{
Output: &out,
InstanceName: instanceName,
Level: slog.LevelInfo,
}).
Info("starting",
"version", "v1")
var fields map[string]any
err := json.Unmarshal(out.Bytes(), &fields)
if err != nil {
t.Fatalf("decode %q: %v", out.String(), err)
}
if fields["type"] != "process" ||
fields["instance"] != "host1/blog" ||
fields["msg"] != "starting" ||
fields["level"] != "INFO" ||
fields["version"] != "v1" {
t.Errorf("process line %v", fields)
}
timeText, _ := fields["time"].(string)
logged, err := time.Parse(time.RFC3339, timeText)
if err != nil || !strings.HasSuffix(timeText, "Z") ||
len(timeText) != len("2006-01-02T15:04:05.000Z") ||
time.Since(logged) > time.Minute {
t.Errorf(
"process line time %q, want now in UTC with milliseconds",
timeText,
)
}
}
func TestProcessLoggerWritesTheMessagesAtItsLevelOrMoreSevere(t *testing.T) {
t.Parallel()
levels := []slog.Level{
slog.LevelDebug, slog.LevelInfo, slog.LevelWarn, slog.LevelError,
}
for i, level := range levels {
t.Run(level.String(), func(t *testing.T) {
t.Parallel()
var out bytes.Buffer
processLog := processlog.New(processlog.Params{
Output: &out,
InstanceName: instanceName,
Level: level,
})
for _, at := range levels {
processLog.Log(t.Context(), at, "message")
}
var got, want []string
for line := range strings.Lines(out.String()) {
var fields struct {
Level string `json:"level"`
}
err := json.Unmarshal([]byte(line), &fields)
if err != nil {
t.Fatalf("decode %q: %v", line, err)
}
got = append(got, fields.Level)
}
for _, written := range levels[i:] {
want = append(want, written.String())
}
if !slices.Equal(got, want) {
t.Errorf("lines at %v, want %v", got, want)
}
})
}
}
+7 -4
View File
@@ -11,7 +11,7 @@ import (
func TestHistoryKeepsEveryRequest(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -57,7 +57,7 @@ func TestHistoryKeepsEveryRequest(t *testing.T) {
func TestLookupReachesTheHistoryOfAClientInTheTable(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
client := netip.MustParsePrefix("203.0.113.9/32")
other := netip.MustParsePrefix("198.51.100.7/32")
start := midnight()
@@ -100,7 +100,10 @@ func TestLookupReachesTheHistoryOfAClientInTheTable(t *testing.T) {
func TestResetKeepsTheHistory(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{PerMinute: limit}, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -123,7 +126,7 @@ func TestResetKeepsTheHistory(t *testing.T) {
func TestRequestsAddsUpTheClientsInsideTheNetblock(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
for client, requests := range map[string]int{
"198.51.100.9/32": 2,
+18 -6
View File
@@ -156,16 +156,28 @@ type Request struct {
BlockedByCoreRuleSet bool
}
// New returns a Limiter for limits, with no client counted yet, whose
// table holds at most maxClients clients (SWW_WAF_MAX_TRACKED_CLIENTS). Past
// it, the least recently seen client is dropped, with its history, and
// starts afresh if it comes back.
func New(limits Limits, maxClients int) *Limiter {
clients, err := simplelru.NewLRU[netip.Prefix, *Client](maxClients, nil)
// Params are what New needs.
type Params struct {
// Limits are what each client's requests and bytes are counted against.
Limits Limits
// MaxClients is the most clients the table holds,
// SWW_WAF_MAX_TRACKED_CLIENTS. Past it, the least recently seen client
// is dropped, with its history, and starts afresh if it comes back.
MaxClients int
}
// New returns a Limiter with no client counted yet.
func New(params Params) *Limiter {
clients, err := simplelru.NewLRU[netip.Prefix, *Client](
params.MaxClients,
nil,
)
if err != nil {
panic(err) // NewLRU fails only for a size below one
}
limits := params.Limits
return &Limiter{
windows: [3]window{
{
+63 -39
View File
@@ -33,10 +33,10 @@ func TestOverARateLimitSaysWhatTheNextRequestWouldDoAndCountsNothing(
client := netip.MustParsePrefix("203.0.113.9/32")
now := midnight()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: limit, PerDay: limit + 1},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit, PerDay: limit + 1},
MaxClients: tableSize,
})
if limiter.OverARateLimit(client, now) {
t.Error("a client never seen is over a rate limit")
@@ -72,7 +72,7 @@ func TestOverARateLimitSaysWhatTheNextRequestWouldDoAndCountsNothing(
wantCount(t, limiter, client, now, minute)
// A limit that is off is never gone over.
off := ratelimit.New(ratelimit.Limits{}, tableSize)
off := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
off.Count(client, now, whole)
if off.OverARateLimit(client, now) {
@@ -95,7 +95,10 @@ func TestEachWindowRefusesAtItsLimitAndLetsTheClientBack(t *testing.T) {
t.Run(tc.window, func(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(tc.limits, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: tc.limits,
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
quarter := tc.length / 4
@@ -126,10 +129,10 @@ func TestEachWindowRefusesAtItsLimitAndLetsTheClientBack(t *testing.T) {
func TestHitGivesTheLimitAndTheRequestsCounted(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: limit, PerHour: limit},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit, PerHour: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -159,10 +162,13 @@ func TestHitGivesTheLimitAndTheRequestsCounted(t *testing.T) {
func TestClientGetsItsPercentageOfEachLimitRoundedDown(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: 5, BytesPerDay: math.MaxInt64},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{
PerMinute: 5,
BytesPerDay: math.MaxInt64,
},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -203,10 +209,10 @@ func TestZeroPercentIsAZeroAllowanceAndALimitOffStaysOff(t *testing.T) {
t.Parallel()
// Only the hour has limits: the minute's and the day's are off.
limiter := ratelimit.New(
ratelimit.Limits{PerHour: limit, BytesPerHour: 1000},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerHour: limit, BytesPerHour: 1000},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -253,7 +259,10 @@ func TestEachByteLimitIsBrokenByTheBytesCounted(t *testing.T) {
t.Run(tc.window, func(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(tc.limits, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: tc.limits,
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
// 600 bytes are within the limit, 600 more over it.
@@ -287,10 +296,10 @@ func TestEachByteLimitIsBrokenByTheBytesCounted(t *testing.T) {
func TestALimitIsBrokenOnlyByWhatIsAddedToIt(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: 2, BytesPerMinute: 1000},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: 2, BytesPerMinute: 1000},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
other := netip.MustParsePrefix("203.0.113.10/32")
start := midnight()
@@ -314,7 +323,7 @@ func TestALimitIsBrokenOnlyByWhatIsAddedToIt(t *testing.T) {
func TestCountGivesTheBytesInEachWindow(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -340,7 +349,10 @@ func TestCountGivesTheBytesInEachWindow(t *testing.T) {
func TestResetSetsTheBytesBackToZero(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{BytesPerDay: 1000}, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{BytesPerDay: 1000},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -356,7 +368,7 @@ func TestRefusalsOverTheLimitInAMinuteBreakTheRefusalLimit(
) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -411,7 +423,7 @@ func TestRefusalsOverTheLimitInAMinuteBreakTheRefusalLimit(
func TestCountGivesTheRequestsInEachWindow(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -438,10 +450,10 @@ func TestCountGivesTheRequestsInEachWindow(t *testing.T) {
func TestResetSetsTheCountsBackToZero(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: limit, PerDay: limit},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit, PerDay: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -463,7 +475,10 @@ func TestResetSetsTheCountsBackToZero(t *testing.T) {
func TestClientBackAfterAWholeBucketIsWithinTheLimitAtOnce(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{PerHour: limit}, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerHour: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -482,10 +497,10 @@ func TestClientBackAfterAWholeBucketIsWithinTheLimitAtOnce(t *testing.T) {
func TestRefusedRequestsCount(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(
ratelimit.Limits{PerMinute: limit, PerHour: 2 * limit},
tableSize,
)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit, PerHour: 2 * limit},
MaxClients: tableSize,
})
refused := netip.MustParsePrefix("203.0.113.9/32")
within := netip.MustParsePrefix("203.0.113.10/32")
start := midnight()
@@ -518,7 +533,10 @@ func TestRefusedRequestsCount(t *testing.T) {
func TestRequestCountedLateGoesInTheBucketUnderWay(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{PerMinute: limit}, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -534,7 +552,10 @@ func TestRequestCountedLateGoesInTheBucketUnderWay(t *testing.T) {
func TestClockSetBackStartsTheBucketsAfresh(t *testing.T) {
t.Parallel()
limiter := ratelimit.New(ratelimit.Limits{PerHour: limit}, tableSize)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerHour: limit},
MaxClients: tableSize,
})
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
@@ -562,7 +583,10 @@ func TestKeepsAtMostMaxClientsDroppingTheLeastRecentlySeen(t *testing.T) {
const maxClients = 3
limiter := ratelimit.New(ratelimit.Limits{PerMinute: 1}, maxClients)
limiter := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerMinute: 1},
MaxClients: maxClients,
})
now := midnight()
clients := make([]netip.Prefix, maxClients+1)
+12 -6
View File
@@ -19,7 +19,7 @@ func TestSnapshotListsTheClientsByAddress(t *testing.T) {
"2001:db8::/64",
}
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
for _, i := range []int{2, 3, 0, 1} {
limiter.Count(netip.MustParsePrefix(want[i]), midnight(), whole)
}
@@ -52,7 +52,10 @@ func TestLoadedCountsCarryOn(t *testing.T) {
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
before := ratelimit.New(ratelimit.Limits{PerHour: limit}, tableSize)
before := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerHour: limit},
MaxClients: tableSize,
})
for range limit {
wantCount(t, before, client, start, "")
}
@@ -60,7 +63,10 @@ func TestLoadedCountsCarryOn(t *testing.T) {
// Loaded into a new limiter, as across a restart, the client has no
// fresh allowance.
later := start.Add(time.Minute)
after := ratelimit.New(ratelimit.Limits{PerHour: limit}, tableSize)
after := ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{PerHour: limit},
MaxClients: tableSize,
})
after.Load(before.Snapshot(), later)
wantCount(t, after, client, later, hour)
}
@@ -71,7 +77,7 @@ func TestLoadEmptiesBucketsWhoseTimeHasPassed(t *testing.T) {
client := netip.MustParsePrefix("203.0.113.9/32")
start := midnight()
limiter := ratelimit.New(ratelimit.Limits{}, tableSize)
limiter := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
limiter.Count(client, start, whole)
limiter.CountBytes(client, start, 5, whole)
limiter.CountRefusal(client, start, limit)
@@ -80,7 +86,7 @@ func TestLoadEmptiesBucketsWhoseTimeHasPassed(t *testing.T) {
loaded := func(now time.Time) ratelimit.Client {
t.Helper()
after := ratelimit.New(ratelimit.Limits{}, tableSize)
after := ratelimit.New(ratelimit.Params{MaxClients: tableSize})
after.Load(limiter.Snapshot(), now)
return after.Snapshot()[0]
@@ -142,7 +148,7 @@ func TestLoadDropsTheLeastRecentlySeenFirst(t *testing.T) {
addr = addr.Next()
}
limiter := ratelimit.New(ratelimit.Limits{}, maxClients)
limiter := ratelimit.New(ratelimit.Params{MaxClients: maxClients})
limiter.Load(clients, midnight())
got := limiter.Snapshot()
+16 -48
View File
@@ -110,12 +110,8 @@ type AbuseIPDB struct {
// scores are by client. Each is added as it is fetched and never moved
// up, so that the one fetched longest ago is the first dropped.
scores *simplelru.LRU[netip.Prefix, Score]
// checking are the clients whose check is under way, each with the
// function that ends it. stopped is true once StopChecking has been
// called, and checkGoroutines waits for the goroutine of each check.
checking map[netip.Prefix]context.CancelFunc
stopped bool
checkGoroutines sync.WaitGroup
// checking are the clients whose check is under way.
checking map[netip.Prefix]bool
// day is the day, in UTC, of the checks spent counts.
day civil.Date
spent int
@@ -138,7 +134,7 @@ func NewAbuseIPDB(params AbuseIPDBParams) *AbuseIPDB {
params: params,
httpClient: &http.Client{},
scores: scores,
checking: map[netip.Prefix]context.CancelFunc{},
checking: map[netip.Prefix]bool{},
}
}
@@ -148,11 +144,11 @@ func NewAbuseIPDB(params AbuseIPDBParams) *AbuseIPDB {
// addresses its request comes from. A client without one is checked in
// the background, by addr, the address its request came from, if
// offender, if it has committed an offence, unless its check is under
// way, a check failed less than failureDelay ago, the day's checks have
// used up DailyBudget, or StopChecking has been called; Hit never waits
// for a check. The check that uses the budget up is logged and raised as a
// source_failure alert. ctx is the context of the client's request, and a
// check goes on after the request ends, until StopChecking.
// way, a check failed less than failureDelay ago, or the day's checks
// have used up DailyBudget; Hit never waits for a check. The check that
// uses the budget up is logged and raised as a source_failure alert. ctx
// is the context of the client's request, and a check goes on after the
// request ends.
func (a *AbuseIPDB) Hit(
ctx context.Context, client netip.Prefix, addr netip.Addr, offender bool,
) (int64, bool) {
@@ -171,21 +167,14 @@ func (a *AbuseIPDB) Hit(
a.day, a.spent = today, 0
}
_, checking := a.checking[client]
check := offender && !checking && !a.stopped &&
!now.Before(a.retryAt) && a.spent < a.params.DailyBudget
check := offender && !a.checking[client] && !now.Before(a.retryAt) &&
a.spent < a.params.DailyBudget
if check {
checkCtx, end := context.WithCancel(context.WithoutCancel(ctx))
a.checking[client] = end
a.checking[client] = true
a.checks++
a.spent++
a.checkGoroutines.Go(func() {
defer end()
a.check(checkCtx, client, addr)
})
go a.check(context.WithoutCancel(ctx), client, addr)
}
usedUp := check && a.spent == a.params.DailyBudget
@@ -200,24 +189,6 @@ func (a *AbuseIPDB) Hit(
return 0, false
}
// StopChecking ends the checks under way, and returns once their goroutines
// have. A check it ends gives no score and is no failure: it is neither
// counted, logged nor raised as an alert, and checks are not left alone
// after it. The ended check stays spent from the day's budget. After
// StopChecking, Hit checks no client.
func (a *AbuseIPDB) StopChecking() {
a.mu.Lock()
a.stopped = true
for _, end := range a.checking {
end()
}
a.mu.Unlock()
a.checkGoroutines.Wait()
}
// Checked returns how many checks were made.
func (a *AbuseIPDB) Checked() int {
a.mu.Lock()
@@ -344,8 +315,7 @@ func (a *AbuseIPDB) RunBlacklist(ctx context.Context) {
// check checks client with AbuseIPDB by addr, one of its addresses, keeps
// the score as client's, and notes the check as no longer under way. A
// check that fails gives no score: it is counted, logged and raised as a
// source_failure alert, and no client is checked for failureDelay. A check
// ended as ctx ends, by StopChecking, is no failure.
// source_failure alert, and no client is checked for failureDelay.
func (a *AbuseIPDB) check(
ctx context.Context,
client netip.Prefix,
@@ -353,26 +323,24 @@ func (a *AbuseIPDB) check(
) {
score, err := a.ask(ctx, addr)
now := a.params.Now()
isFailure := err != nil && ctx.Err() == nil
a.mu.Lock()
delete(a.checking, client)
switch {
case err == nil:
if err == nil {
a.scores.Add(
client,
Score{Client: client, Score: score, Fetched: now},
)
case isFailure:
} else {
a.failures++
a.retryAt = now.Add(failureDelay)
}
a.mu.Unlock()
if isFailure {
if err != nil {
a.alert("checking a client with AbuseIPDB failed", err)
}
}
-37
View File
@@ -394,43 +394,6 @@ func TestCheckNotAnsweredWithinTheTimeoutFailsAndHitNeverWaits(
})
}
func TestStopCheckingEndsTheCheckUnderWayAsNoFailureAndChecksNoMore(
t *testing.T,
) {
t.Parallel()
synctest.Test(t, func(t *testing.T) {
queue := newQueue()
p := abuseIPDBParams()
p.Alerts = queue
abuseIPDB := &abuseIPDBStandIn{hanging: true}
checker := newAbuseIPDB(abuseIPDB, p)
wantScore(t, checker, suspect, true, 0, false)
synctest.Wait()
// The check is ended, not waited out, and is no failure.
began := time.Now()
checker.StopChecking()
if waited := time.Since(began); waited != 0 {
t.Errorf("StopChecking returned after %s, want at once", waited)
}
wantFailures(t, checker, 0)
if got := waiting(queue); len(got) != 0 {
t.Errorf("alerts waiting %+v, want none", got)
}
// An offender without a score is then not checked.
wantScore(t, checker, other, true, 0, false)
synctest.Wait()
wantChecked(t, abuseIPDB, suspect)
})
}
func TestKeyIsSentInTheKeyHeaderAndNeverShown(t *testing.T) {
t.Parallel()
+13 -44
View File
@@ -82,12 +82,8 @@ type DNSBL struct {
// verdicts are by query. Each is added as it is fetched and never moved
// up, so that the one fetched longest ago is the first dropped.
verdicts *simplelru.LRU[query, Verdict]
// asking are the queries under way, each with the function that ends
// it. stopped is true once StopAsking has been called, and
// queryGoroutines waits for the goroutine of each query.
asking map[query]context.CancelFunc
stopped bool
queryGoroutines sync.WaitGroup
// asking are the queries under way.
asking map[query]bool
// queries and failures count, by zone, the queries made and those that
// failed, and retryAt is when a zone whose last query failed may be
// asked again.
@@ -124,7 +120,7 @@ func NewDNSBL(params DNSBLParams) *DNSBL {
params: params,
resolver: resolver,
verdicts: verdicts,
asking: map[query]context.CancelFunc{},
asking: map[query]bool{},
queries: map[string]int{},
failures: map[string]int{},
retryAt: map[string]time.Time{},
@@ -142,10 +138,9 @@ func (d *DNSBL) Zones() []string {
// log, the alerts and the metrics. A verdict is used until CacheTTL has
// passed since it was fetched. Each zone without one is asked about addr
// in the background, unless a query about addr to it is under way, the
// zone is left alone after a failure, maxQueries are under way, or
// StopAsking has been called; ListedBy never waits for a query. ctx is the
// context of the client's request, and a query goes on after the request
// ends, until StopAsking.
// zone is left alone after a failure, or maxQueries are under way;
// ListedBy never waits for a query. ctx is the context of the client's
// request, and a query goes on after the request ends.
func (d *DNSBL) ListedBy(ctx context.Context, addr netip.Addr) []string {
d.mu.Lock()
defer d.mu.Unlock()
@@ -158,47 +153,23 @@ func (d *DNSBL) ListedBy(ctx context.Context, addr netip.Addr) []string {
q := query{zone: zone, client: addr}
kept, found := d.verdicts.Peek(q)
_, asking := d.asking[q]
switch {
case found && now.Sub(kept.Fetched) < d.params.CacheTTL:
if kept.Listed {
listedBy = append(listedBy, config.MaskZoneKey(zone))
}
case !asking && !d.stopped && !now.Before(d.retryAt[zone]) &&
len(d.asking) < maxQueries:
queryCtx, end := context.WithCancel(context.WithoutCancel(ctx))
d.asking[q] = end
case !d.asking[q] && !now.Before(d.retryAt[zone]) && len(d.asking) < maxQueries:
d.asking[q] = true
d.queries[zone]++
d.queryGoroutines.Go(func() {
defer end()
d.ask(queryCtx, q)
})
go d.ask(context.WithoutCancel(ctx), q)
}
}
return listedBy
}
// StopAsking ends the queries under way, and returns once their
// goroutines have. A query it ends gives no verdict and is no failure: it
// is neither counted, logged nor raised as an alert, and the zone is not
// left alone after it. After StopAsking, ListedBy asks no zone.
func (d *DNSBL) StopAsking() {
d.mu.Lock()
d.stopped = true
for _, end := range d.asking {
end()
}
d.mu.Unlock()
d.queryGoroutines.Wait()
}
// Queries returns how many queries were made to zone.
func (d *DNSBL) Queries(zone string) int {
d.mu.Lock()
@@ -267,29 +238,27 @@ func (d *DNSBL) Load(verdicts []Verdict) {
// query as no longer under way. A query that fails gives no verdict: it
// is counted, logged and raised as a source_failure alert, which show the
// zone with its key masked, and the zone is not asked again for
// failureDelay. A query ended as ctx ends, by StopAsking, is no failure.
// failureDelay.
func (d *DNSBL) ask(ctx context.Context, q query) {
listed, err := d.lookUp(ctx, q)
now := d.params.Now()
isFailure := err != nil && ctx.Err() == nil
d.mu.Lock()
delete(d.asking, q)
switch {
case err == nil:
if err == nil {
d.verdicts.Add(q, Verdict{
Zone: q.zone, Client: q.client, Listed: listed, Fetched: now,
})
case isFailure:
} else {
d.failures[q.zone]++
d.retryAt[q.zone] = now.Add(failureDelay)
}
d.mu.Unlock()
if isFailure {
if err != nil {
const failed = "asking a DNSBL zone failed"
shown := config.MaskZoneKey(q.zone)
-39
View File
@@ -210,45 +210,6 @@ func TestQueryNotAnsweredWithinTheTimeoutFails(t *testing.T) {
})
}
//
//nolint:paralleltest // one at a time, as the comment at the top of this file says
func TestStopAskingEndsTheQueryUnderWayAsNoFailureAndAsksNoMore(
t *testing.T,
) {
synctest.Test(t, func(t *testing.T) {
queue := newQueue()
p := dnsblParams(zone)
p.Alerts = queue
resolver := &resolverStandIn{hanging: true}
dnsbl := newDNSBL(resolver, p)
wantZones(t, dnsbl, listed)
synctest.Wait()
// The query is ended, not waited out, and is no failure.
began := time.Now()
dnsbl.StopAsking()
if waited := time.Since(began); waited != 0 {
t.Errorf("StopAsking returned after %s, want at once", waited)
}
wantQueries(t, dnsbl, 1, 0)
if got := waiting(queue); len(got) != 0 {
t.Errorf("alerts waiting %+v, want none", got)
}
// A client without a verdict is then asked about in no zone.
wantZones(t, dnsbl, unlisted)
synctest.Wait()
wantQueries(t, dnsbl, 1, 0)
wantAsked(t, resolver, listedName)
waitForTheResolver()
})
}
//
//nolint:paralleltest // one at a time, as the comment at the top of this file says
func TestZoneThatFailsOrRefusesGivesNoVerdictAndIsLeftAloneForAMinute(
+3 -32
View File
@@ -1,7 +1,7 @@
// Package requestlog writes the lines sww prints on stdout: one
// Package requestlog writes the request log sww prints on stdout: one
// JSON object per request, two from waf and a short one from rproxy,
// marked "type":"request", and the process's own messages as JSON lines marked
// "type":"process".
// marked "type":"request". Package processlog writes the process's own
// messages.
package requestlog
import (
@@ -9,7 +9,6 @@ import (
"encoding/json"
"fmt"
"io"
"log/slog"
"time"
"sneak.berlin/go/smallwebwaf/internal/ratelimit"
@@ -300,31 +299,3 @@ func Milliseconds(d time.Duration) float64 {
time.Millisecond/time.Microsecond,
)
}
// NewProcessLogger returns the logger for the process's own messages:
// JSON lines on w, marked "type":"process", with the time in the same form
// as a request line's, and instanceName, SWW_WAF_INSTANCE_NAME, as instance.
// It writes only the messages at level, SWW_WAF_LOG_LEVEL, or more severe;
// the request lines Write writes are never held back.
func NewProcessLogger(
w io.Writer,
instanceName string,
level slog.Level,
) *slog.Logger {
handler := slog.NewJSONHandler(w, &slog.HandlerOptions{
Level: level,
ReplaceAttr: func(groups []string, attr slog.Attr) slog.Attr {
if attr.Key == slog.TimeKey && len(groups) == 0 {
return slog.String(
slog.TimeKey,
FormatTime(attr.Value.Time()),
)
}
return attr
},
})
return slog.New(handler).
With("type", "process", "instance", instanceName)
}
-87
View File
@@ -3,8 +3,6 @@ package requestlog_test
import (
"bytes"
"encoding/json"
"log/slog"
"slices"
"strings"
"testing"
"time"
@@ -134,88 +132,3 @@ func TestWriteRProxyLineWritesOneJSONLineWithOnlyItsFields(t *testing.T) {
t.Errorf("wrote %s, want %s", out.String(), want)
}
}
func TestProcessLinesAreMarkedProcessAndGiveTheInstance(t *testing.T) {
t.Parallel()
var out bytes.Buffer
requestlog.NewProcessLogger(&out, "host1/blog", slog.LevelInfo).
Info("starting",
"version", "v1")
var fields map[string]any
err := json.Unmarshal(out.Bytes(), &fields)
if err != nil {
t.Fatalf("decode %q: %v", out.String(), err)
}
if fields["type"] != "process" ||
fields["instance"] != "host1/blog" ||
fields["msg"] != "starting" ||
fields["level"] != "INFO" ||
fields["version"] != "v1" {
t.Errorf("process line %v", fields)
}
timeText, _ := fields["time"].(string)
logged, err := time.Parse(time.RFC3339, timeText)
if err != nil || !strings.HasSuffix(timeText, "Z") ||
len(timeText) != len("2006-01-02T15:04:05.000Z") ||
time.Since(logged) > time.Minute {
t.Errorf(
"process line time %q, want now in UTC with milliseconds",
timeText,
)
}
}
func TestProcessLoggerWritesTheMessagesAtItsLevelOrMoreSevere(t *testing.T) {
t.Parallel()
levels := []slog.Level{
slog.LevelDebug, slog.LevelInfo, slog.LevelWarn, slog.LevelError,
}
for i, level := range levels {
t.Run(level.String(), func(t *testing.T) {
t.Parallel()
var out bytes.Buffer
processLog := requestlog.NewProcessLogger(
&out,
"host1/blog",
level,
)
for _, at := range levels {
processLog.Log(t.Context(), at, "message")
}
var got, want []string
for line := range strings.Lines(out.String()) {
var fields struct {
Level string `json:"level"`
}
err := json.Unmarshal([]byte(line), &fields)
if err != nil {
t.Fatalf("decode %q: %v", line, err)
}
got = append(got, fields.Level)
}
for _, written := range levels[i:] {
want = append(want, written.String())
}
if !slices.Equal(got, want) {
t.Errorf("lines at %v, want %v", got, want)
}
})
}
}
+12 -43
View File
@@ -99,12 +99,8 @@ type Client struct {
// answers are by client. Each is added as it comes and never moved up,
// so that the one that came longest ago is the first dropped.
answers *simplelru.LRU[netip.Prefix, Answer]
// asking are the clients whose query is under way, each with the
// function that ends it. stopped is true once StopAsking has been
// called, and queryGoroutines waits for the goroutine of each query.
asking map[netip.Prefix]context.CancelFunc
stopped bool
queryGoroutines sync.WaitGroup
// asking are the clients whose query is under way.
asking map[netip.Prefix]bool
// retryAt is when routewatch may be asked again after a query failed.
retryAt time.Time
}
@@ -120,7 +116,7 @@ func New(params Params) *Client {
params: params,
httpClient: &http.Client{},
answers: answers,
asking: map[netip.Prefix]context.CancelFunc{},
asking: map[netip.Prefix]bool{},
}
}
@@ -128,9 +124,8 @@ func New(params Params) *Client {
// IPv6 group, or why it has no route. It never waits for a query. A client
// without an answer is asked about in the background, by addr, the address its
// request came from, unless its query is under way, a query failed less than
// failureKept ago, maxQueries are under way, or StopAsking has been called. ctx
// is the context of the client's request, and a query goes on after the request
// ends, until StopAsking.
// failureKept ago, or maxQueries are under way. ctx is the context of the
// client's request, and a query goes on after the request ends.
func (c *Client) Route(
ctx context.Context, client netip.Prefix, addr netip.Addr,
) (Route, Missing) {
@@ -152,38 +147,15 @@ func (c *Client) Route(
return Route{}, RequestFailed
}
_, asking := c.asking[client]
if !asking && !c.stopped && len(c.asking) < maxQueries {
queryCtx, end := context.WithCancel(context.WithoutCancel(ctx))
c.asking[client] = end
if !c.asking[client] && len(c.asking) < maxQueries {
c.asking[client] = true
c.queryGoroutines.Go(func() {
defer end()
c.ask(queryCtx, client, addr)
})
go c.ask(context.WithoutCancel(ctx), client, addr)
}
return Route{}, NoAnswerYet
}
// StopAsking ends the queries under way, and returns once their goroutines
// have. A query it ends gives no answer and is no failure: it is not logged,
// and routewatch is not left alone after it. After StopAsking, Route asks
// routewatch nothing.
func (c *Client) StopAsking() {
c.mu.Lock()
c.stopped = true
for _, end := range c.asking {
end()
}
c.mu.Unlock()
c.queryGoroutines.Wait()
}
// Snapshot returns every answer still in use, sorted by client, as
// routewatch.json lists them.
func (c *Client) Snapshot() []Answer {
@@ -233,8 +205,7 @@ func (c *Client) Load(answers []Answer) {
// ask asks routewatch about addr, keeps the answer as client's, and notes the
// query as no longer under way. A query that fails gives no answer, and
// routewatch is not asked again for failureKept. Only the first failure of
// those under way is logged. A query ended as ctx ends, by StopAsking, is no
// failure.
// those under way is logged.
func (c *Client) ask(
ctx context.Context,
client netip.Prefix,
@@ -242,20 +213,18 @@ func (c *Client) ask(
) {
route, err := c.query(ctx, addr)
now := c.params.Now()
isFailure := err != nil && ctx.Err() == nil
c.mu.Lock()
delete(c.asking, client)
firstFailure := isFailure && !now.Before(c.retryAt)
firstFailure := err != nil && !now.Before(c.retryAt)
switch {
case err == nil:
if err == nil {
c.answers.Add(client, Answer{
Client: client, Netblock: route.Netblock, ASN: route.ASN, Answered: now,
})
case isFailure:
} else {
c.retryAt = now.Add(failureKept)
}
-48
View File
@@ -361,54 +361,6 @@ func TestHangingRoutewatchNeverDelaysAndFailsTheQuery(t *testing.T) {
})
}
func TestStopAskingEndsTheQueryUnderWayAsNoFailureAndAsksNoMore(
t *testing.T,
) {
t.Parallel()
// In a synctest bubble, as TestHangingRoutewatchNeverDelaysAndFailsTheQuery
// says.
synctest.Test(t, func(t *testing.T) {
f := &fixture{t: t, log: &syncBuffer{}}
f.client = routewatch.New(routewatch.Params{
URL: silentURL, Now: time.Now,
ProcessLog: slog.New(slog.NewTextHandler(f.log, nil)),
})
f.route()
synctest.Wait()
// The query is ended, not waited out.
began := time.Now()
f.client.StopAsking()
if waited := time.Since(began); waited != 0 {
t.Errorf("StopAsking returned after %s, want at once", waited)
}
// A query started now would fail within waitLimit. None is, and the
// one ended is no failure: nothing is logged, and routewatch is not
// left alone.
for range 2 {
if _, missing := f.route(); missing != routewatch.NoAnswerYet {
t.Errorf(
"Route gave %q, want %q",
missing,
routewatch.NoAnswerYet,
)
}
time.Sleep(waitLimit)
synctest.Wait()
}
if f.log.String() != "" {
t.Errorf("process log %q, want nothing", f.log.String())
}
})
}
func TestRepeatedRequestsMakeOneQueryAndTheAnswerExpires(t *testing.T) {
t.Parallel()
+6 -6
View File
@@ -18,7 +18,7 @@ import (
"sneak.berlin/go/smallwebwaf/internal/config"
"sneak.berlin/go/smallwebwaf/internal/config/configtest"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
"sneak.berlin/go/smallwebwaf/internal/processlog"
"sneak.berlin/go/smallwebwaf/internal/rproxy"
"sneak.berlin/go/smallwebwaf/internal/waf"
)
@@ -51,11 +51,11 @@ func TestServerHasTheFixedLimits(t *testing.T) {
server := rproxy.New(rproxy.Params{
Config: cfg,
RequestLog: io.Discard,
ProcessLog: requestlog.NewProcessLogger(
io.Discard,
"test",
slog.LevelInfo,
),
ProcessLog: processlog.New(processlog.Params{
Output: io.Discard,
InstanceName: "test",
Level: slog.LevelInfo,
}),
})
if server.Addr != ":8080" || server.MaxHeaderBytes != 28<<10 ||
+6 -6
View File
@@ -21,7 +21,7 @@ import (
"sneak.berlin/go/smallwebwaf/internal/config"
"sneak.berlin/go/smallwebwaf/internal/config/configtest"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
"sneak.berlin/go/smallwebwaf/internal/processlog"
"sneak.berlin/go/smallwebwaf/internal/rproxy"
"sneak.berlin/go/smallwebwaf/internal/waf"
)
@@ -370,11 +370,11 @@ func startServerLogging(
server := rproxy.New(rproxy.Params{
Config: cfg,
RequestLog: requestLog,
ProcessLog: requestlog.NewProcessLogger(
processLog,
"test",
slog.LevelInfo,
),
ProcessLog: processlog.New(processlog.Params{
Output: processLog,
InstanceName: "test",
Level: slog.LevelInfo,
}),
})
listener, err := (&net.ListenConfig{}).Listen(
+1 -1
View File
@@ -2448,7 +2448,7 @@ func newParams(dir string) state.Params {
AttackBanDuration: 7 * 24 * time.Hour,
MaxBans: 5000,
}),
Limiter: ratelimit.New(ratelimit.Limits{}, 20000),
Limiter: ratelimit.New(ratelimit.Params{MaxClients: 20000}),
GeoJS: geojs.New(geojs.Params{Now: midnight}),
Lists: reputation.New(reputation.Params{
BlocklistURLs: []string{
+6 -6
View File
@@ -10,7 +10,7 @@ import (
"go.uber.org/fx"
"sneak.berlin/go/smallwebwaf/internal/config"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
"sneak.berlin/go/smallwebwaf/internal/processlog"
"sneak.berlin/go/smallwebwaf/internal/rproxy"
)
@@ -23,11 +23,11 @@ import (
// setting for either.
func RunRProxy(ctx context.Context, params Params) int {
hostname, _ := os.Hostname() // "" when the host has no name to give
processLog := requestlog.NewProcessLogger(
params.Stdout,
hostname,
slog.LevelInfo,
)
processLog := processlog.New(processlog.Params{
Output: params.Stdout,
InstanceName: hostname,
Level: slog.LevelInfo,
})
cfg, err := config.RProxyFromEnvironment(params.Viper)
if err != nil {
+34 -52
View File
@@ -19,10 +19,10 @@ import (
"sneak.berlin/go/smallwebwaf/internal/geojs"
"sneak.berlin/go/smallwebwaf/internal/lookup"
"sneak.berlin/go/smallwebwaf/internal/metrics"
"sneak.berlin/go/smallwebwaf/internal/processlog"
"sneak.berlin/go/smallwebwaf/internal/ratelimit"
"sneak.berlin/go/smallwebwaf/internal/remotelog"
"sneak.berlin/go/smallwebwaf/internal/reputation"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
"sneak.berlin/go/smallwebwaf/internal/routewatch"
"sneak.berlin/go/smallwebwaf/internal/rules"
"sneak.berlin/go/smallwebwaf/internal/state"
@@ -43,8 +43,11 @@ const remoteLogStopTimeout = 2 * time.Second
func RunWAF(ctx context.Context, params Params) int {
// Until the settings are read, the one message is an invalid setting's
// error, which every SWW_WAF_LOG_LEVEL lets through.
processLog := requestlog.NewProcessLogger(params.Stdout,
config.InstanceName(params.Viper), slog.LevelError)
processLog := processlog.New(processlog.Params{
Output: params.Stdout,
InstanceName: config.InstanceName(params.Viper),
Level: slog.LevelError,
})
cfg, err := config.WAFFromEnvironment(params.Viper)
if err != nil {
@@ -63,11 +66,11 @@ func RunWAF(ctx context.Context, params Params) int {
stdout = io.MultiWriter(params.Stdout, remote)
}
processLog = requestlog.NewProcessLogger(
stdout,
cfg.InstanceName,
cfg.LogLevel,
)
processLog = processlog.New(processlog.Params{
Output: stdout,
InstanceName: cfg.InstanceName,
Level: cfg.LogLevel,
})
if remote != nil {
stopSending := startSending(ctx, remote, processLog)
@@ -118,8 +121,7 @@ func RunWAF(ctx context.Context, params Params) int {
"settings", cfg)
// The last write follows the stop of every hook, and so the end of every
// loop that changes the state or raises an alert, and of every ask of
// GeoJS, the DNSBL zones, AbuseIPDB and routewatch. A hook of the state
// loop that changes the state or raises an alert. A hook of the state
// files would stop before the loops of the parts built ahead of them.
return run(ctx, app, failed, processLog, func() error {
err := files.WriteAll()
@@ -239,14 +241,17 @@ func newLookupFile(
// newLimiter returns the table of clients, with the rate and byte limits.
func newLimiter(cfg *config.WAF) *ratelimit.Limiter {
return ratelimit.New(ratelimit.Limits{
PerMinute: cfg.RateLimitPerMinute,
PerHour: cfg.RateLimitPerHour,
PerDay: cfg.RateLimitPerDay,
BytesPerMinute: cfg.BytesLimitPerMinute,
BytesPerHour: cfg.BytesLimitPerHour,
BytesPerDay: cfg.BytesLimitPerDay,
}, cfg.MaxTrackedClients)
return ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{
PerMinute: cfg.RateLimitPerMinute,
PerHour: cfg.RateLimitPerHour,
PerDay: cfg.RateLimitPerDay,
BytesPerMinute: cfg.BytesLimitPerMinute,
BytesPerHour: cfg.BytesLimitPerHour,
BytesPerDay: cfg.BytesLimitPerDay,
},
MaxClients: cfg.MaxTrackedClients,
})
}
// newLedger returns the ledger of the bans, with the settings for them.
@@ -298,13 +303,11 @@ func newLists(
return lists
}
// newDNSBL returns what asks the DNSBL zones, none asked yet. At the stop it
// ends the queries under way, as reputation.DNSBL.StopAsking says.
// newDNSBL returns what asks the DNSBL zones, none asked yet.
func newDNSBL(
lc fx.Lifecycle, cfg *config.WAF, processLog *slog.Logger,
alertQueue *alerts.Queue,
cfg *config.WAF, processLog *slog.Logger, alertQueue *alerts.Queue,
) *reputation.DNSBL {
dnsbl := reputation.NewDNSBL(reputation.DNSBLParams{
return reputation.NewDNSBL(reputation.DNSBLParams{
Zones: cfg.DNSBLZones,
Resolver: cfg.DNSBLResolver,
CacheTTL: cfg.IPReputationCacheTTL,
@@ -313,17 +316,12 @@ func newDNSBL(
ProcessLog: processLog,
Alerts: alertQueue,
})
lc.Append(fx.StopHook(dnsbl.StopAsking))
return dnsbl
}
// newAbuseIPDB returns what checks clients with AbuseIPDB while
// SWW_WAF_ABUSEIPDB_KEY is set, none checked yet. Until waf stops, it
// fetches AbuseIPDB's blacklist as it is due, and bans the clients on it in
// ledger, as waf.BanAbuseIPDBBlacklist says. At the stop it ends the checks
// under way, as reputation.AbuseIPDB.StopChecking says.
// ledger, as waf.BanAbuseIPDBBlacklist says.
func newAbuseIPDB(
lc fx.Lifecycle, cfg *config.WAF, processLog *slog.Logger,
alertQueue *alerts.Queue, ledger *bans.Ledger,
@@ -349,30 +347,24 @@ func newAbuseIPDB(
})
lc.Append(background(abuseIPDB.RunBlacklist))
lc.Append(fx.StopHook(abuseIPDB.StopChecking))
return abuseIPDB
}
// newRouteWatch returns what asks routewatch about clients, or nil while
// SWW_WAF_ROUTEWATCH_URL is unset. At the stop it ends the queries under
// way, as routewatch.Client.StopAsking says.
// SWW_WAF_ROUTEWATCH_URL is unset.
func newRouteWatch(
lc fx.Lifecycle, cfg *config.WAF, processLog *slog.Logger,
cfg *config.WAF, processLog *slog.Logger,
) *routewatch.Client {
if cfg.RouteWatchURL == "" {
return nil
}
routeWatch := routewatch.New(routewatch.Params{
return routewatch.New(routewatch.Params{
URL: cfg.RouteWatchURL,
Now: utcNow,
ProcessLog: processLog,
})
lc.Append(fx.StopHook(routeWatch.StopAsking))
return routeWatch
}
// newMetrics returns the metrics, with those read from the other parts as
@@ -410,18 +402,12 @@ func newMetrics(
// newGeoJS returns what looks clients up at GeoJS, none looked up yet. Each
// answer goes to the limiter and the ledger, as waf.AddLookup says, and each
// failure is handled as waf.GeoJSFailed says. At the stop it ends the request
// under way, as geojs.GeoJS.StopAsking says.
// failure is handled as waf.GeoJSFailed says.
func newGeoJS(
lc fx.Lifecycle,
cfg *config.WAF,
m *metrics.Metrics,
limiter *ratelimit.Limiter,
ledger *bans.Ledger,
alertQueue *alerts.Queue,
processLog *slog.Logger,
cfg *config.WAF, m *metrics.Metrics, limiter *ratelimit.Limiter,
ledger *bans.Ledger, alertQueue *alerts.Queue, processLog *slog.Logger,
) *geojs.GeoJS {
geoJS := geojs.New(geojs.Params{
return geojs.New(geojs.Params{
URL: geojs.URL,
Answered: func(answer geojs.Answer) {
waf.AddLookup(cfg, limiter, ledger, answer)
@@ -433,10 +419,6 @@ func newGeoJS(
},
WentWithoutAnswer: m.RequestsWithoutGeoJSAnswer.Inc,
})
lc.Append(fx.StopHook(geoJS.StopAsking))
return geoJS
}
// newWAFServer returns the server waf runs, with its parts and the time
+22 -18
View File
@@ -24,6 +24,7 @@ import (
"sneak.berlin/go/smallwebwaf/internal/geojs"
"sneak.berlin/go/smallwebwaf/internal/lookup"
"sneak.berlin/go/smallwebwaf/internal/metrics"
"sneak.berlin/go/smallwebwaf/internal/processlog"
"sneak.berlin/go/smallwebwaf/internal/ratelimit"
"sneak.berlin/go/smallwebwaf/internal/reputation"
"sneak.berlin/go/smallwebwaf/internal/requestlog"
@@ -398,11 +399,11 @@ func startRProxy(
server := rproxy.New(rproxy.Params{
Config: cfg,
RequestLog: io.Discard,
ProcessLog: requestlog.NewProcessLogger(
io.Discard,
"rproxy",
slog.LevelInfo,
),
ProcessLog: processlog.New(processlog.Params{
Output: io.Discard,
InstanceName: "rproxy",
Level: slog.LevelInfo,
}),
})
listener, err := (&net.ListenConfig{}).Listen(
@@ -471,11 +472,11 @@ func newWAFWith(
}
out := &output{}
processLog := requestlog.NewProcessLogger(
out,
cfg.InstanceName,
cfg.LogLevel,
)
processLog := processlog.New(processlog.Params{
Output: out,
InstanceName: cfg.InstanceName,
Level: cfg.LogLevel,
})
ruleFiles, err := rules.Load(rules.Params{
Dir: cfg.RuleFilesDir, Enabled: cfg.RuleFilesEnabled, ProcessLog: processLog,
@@ -561,14 +562,17 @@ func addCounters(params *waf.Params) {
cfg.MetricsBusiestCountriesAndASNs,
cfg.InstanceName,
)
params.Limiter = ratelimit.New(ratelimit.Limits{
PerMinute: cfg.RateLimitPerMinute,
PerHour: cfg.RateLimitPerHour,
PerDay: cfg.RateLimitPerDay,
BytesPerMinute: cfg.BytesLimitPerMinute,
BytesPerHour: cfg.BytesLimitPerHour,
BytesPerDay: cfg.BytesLimitPerDay,
}, cfg.MaxTrackedClients)
params.Limiter = ratelimit.New(ratelimit.Params{
Limits: ratelimit.Limits{
PerMinute: cfg.RateLimitPerMinute,
PerHour: cfg.RateLimitPerHour,
PerDay: cfg.RateLimitPerDay,
BytesPerMinute: cfg.BytesLimitPerMinute,
BytesPerHour: cfg.BytesLimitPerHour,
BytesPerDay: cfg.BytesLimitPerDay,
},
MaxClients: cfg.MaxTrackedClients,
})
params.Ledger = bans.New(bans.Params{
LimitBanDuration: cfg.LimitBanDuration,
LimitBanRepeatWindow: cfg.LimitBanRepeatWindow,